INDEX

Explanations

negations and exclusions in mathematical contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ंदीखरीदारी

-0.57

ValueStyle

-0.56

 nahilalakip

-0.56

niña

-0.52

colhead

-0.49

Aholisi

-0.49

EndProject

-0.49

Personensuche

-0.49

 StatefulWidget

-0.48

 Accesorios

-0.47

POSITIVE LOGITS

RTLR

0.40

 betweenstory

0.31

 rosen

0.31

AV

0.30

rowse

0.29

leg

0.28

LEn

0.28

rig

0.28

gné

0.28

 来自

0.28

Activations Density 0.201%

negations and exclusions in mathematical contexts

No Comments

No Known Activations